3D plug track - version 244c and beyond

[KHC1]

Thanks again Martin and Steve. I have just trialled a B7 and have managed to get a full turnout split and sliced for my printer. I will print Tomorrows as I have a filing jig printing at the moment.

Keith

 

[Steve_Cornford]

Hi Keith,
thats good news. dont forget to change your cura profile from filing jigs back to timbering bricks (I did first time)
Steve

 

[Steve_Cornford]

Hi Martin,
Although I have probably missed something, I seem to have missing connector clips on export having reloaded a box file and the corresponding shapes file from a previous attempt at producing a Barry Slip.
I have three different colour bricks with paired clips between them. On trying to export my middle section it omits the clips.

I have attached the box and shapes files.
When I inspect background shapes the clip colours appear correct, and the are annotated as clips.

Steve

 

[Martin Wynne]

Hi Martin,
Although I have probably missed something, I seem to have missing connector clips on export having reloaded a box file and the corresponding shapes file from a previous attempt at producing a Barry Slip.
I have three different colour bricks with paired clips between them. On trying to export my middle section it omits the clips.
View attachment 9749
I have attached the box and shapes files.
When I inspect background shapes the clip colours appear correct, and the are annotated as clips.
View attachment 9750
Steve

@Steve_Cornford

Hi Steve,

It's working ok here:

Have you got the brick option selected and the correct colour:

If that's ok, check on the layers tab that the background shapes combo isn't blank (it can be any colour).

If all is ok, I need to delve a bit deeper.

cheers,

Martin.

 

[Steve_Cornford]

Hi Martin,
Yes I have the brick option selected.
But I have backgroundshapes blank in the layers tab, but do have the brick connector clips set to green (the default)

I have now set the background shapes to green, and the clips are now exported.
Thank you.
ps However I did not knowingly set it to spaces.
I did perform a reset all , then timbers only -laser 2D, before then timbers only -FDM 3d.
Steve

 

[Steve_Cornford]

Hi Martin,
just repeated the steps, reset all gives:-

Then timbers only - laser 2D gives:-

and then Timbers only -FDM 3D gives:-

which re-instates brick connect clips to green, but leaves background shapes blank.

I was just using the timbers only - laser 2D to more easily check the socket positions on the timbers.
Sorry about that.
Steve

 

[Martin Wynne]

@Steve_Cornford

Hi Steve,

Thanks for finding that bug. It does illustrate how time-consuming it can be to test everything.

timbers only FDM should be switching the background shapes layer back on. I will get it fixed -- or maybe add yet another tick-box option to choose whether you want the brick assemblers included with FDM timbers.

The underlying issue is that I intended making use of the background shapes for the brick assemblers to be a temporary kludge. The assemblers need a separate function of their own, but I suspect it will be a while before I do anything about that.

Thanks again,

Martin.

 

[Martin Wynne]

@nickom @NoIdea @Rusty @Alistair Ward @graeme @Steve_Cornford @James Walters @Phil G

One reason for pressing on with getting the current files released open-source was the hope that someone might be able to help with coding some of the missing chairs.

Because I've realised that for me to do them all myself is going to take a very long time. The REA K-crossing* chairs are long overdue, and I have barely looked at them yet. Then the full list of REA heavable plain-rail chairs is still needed. And a full range of GWR chairs, and other pre-grouping designs. Not to mention flat-bottom baseplates ...

There have been several kind offers of help in creating CAD-style STL files of various chairs. Unfortunately that wouldn't be much help by itself, because Templot creates the chairs programmatically -- to match the required rail location, crossing angle, flangeway gap, model rail section, wheel flange depth clearance, switch opening, increased check rail gap for gauge-widening, etc. I would need to extract a lot of dimensional data from such an STL file and my CAD skills are not good enough to do that.

But if you are good at CAD, or at least better than me, there is maybe a way you could help.

The chairs are divided into several component parts, so it is necessary to do only one part at a time:

chair base up to an imaginary "plinth" level. This is a simple 8-sided 2D outline with radiused corners and sloping sides.

seat block to support the rail. Thicker than the chair base.

outer jaw which sits on the plinth against the seat block.

key for outer jaw which fits between the jaw and the rail web.

inner jaw which sits on the plinth and is in two parts:

the "stand" part which fits against the rail foot.​

the "grip" part which fits between the stand and the rail web.​

several screw/bolt bosses and heads.

Some of these are common across the different chair sizes and designs, so need doing only once.

After creating one of these, what I would need is a series of horizontal cross sections at different levels. Similar to those produced by the slicer (but at wider spacing, say every half-inch). For an outer jaw with 2 ribs such a cross-section might look something like this:

And then what I really need is a list of X,Y co-ordinates for nodes along the perimeter, as I have started indicating with red dots. Ideally evenly spaced, and importantly with the same number of nodes on each layer.

Now I believe I could extract such a list from an STL file using CAD software, but I really don't know how to do such a thing. But if you are good at CAD maybe you could do it? Design the jaw and then slice it up, and create a polyline along the perimeter with a defined number of nodes?

Even if you could send me only the outline shape image, I can write a graphics utility to get the co-ordinates using edge-detection on the image dots.

Armed with such lists of co-ordinates I can write the chair-creation code quite quickly. It is getting the co-ordinate data from the prototype drawing which takes the time.

Thanks for any help you can offer in this direction.

---

*there is a feasible temporary kludge for K-crossings -- video clip at: https://85a.uk/templot/club/index.php?posts/10760

---

Martin.

 

[James Walters]

Hi Martin,
Just seen your post on the Scalefour forum. What you have described above is tedious and dull work which you shouldn't be doing. It's a waste of a creative mind.
I'm certain it can be done using ChatGPT. However, I don't know how to do it, but I know a guy who does do it professionally. I'm seeing him at the weekend so I'll pick his brain.
Until then, tedious and dull CAD operations are up my street, and I'd cheerfully have a go at doing it manually as you describe above. I'm guessing (not assuming) that maybe a spreadsheet of node coordinates X.Y,Z would be useful along with a 3d model which can be visually inspected, and a plot of the sections would be helpful too.
Extracting cross-sections from Solidworks is very simple, so if you can write a code to automate plotting the nodes then this could be a very good way forward.
I can forsee the need to use parametric variables within the drawing for things such as rail section, loose-jaw slot etc, but it would be interesting to have a go.

If I've not misunderstood the task, would you like to suggest a particular chair for me to have a look at?

This would be an excellent way for me to fill-time and look busy when in meetings at work. I've completed the levels in my sudoko app so need a fresh challenge.

All the best,
James

 

[Martin Wynne]

Until then, tedious and dull CAD operations are up my street, and I'd cheerfully have a go at doing it manually as you describe above. I'm guessing (not assuming) that maybe a spreadsheet of node coordinates X.Y,Z would be useful along with a 3d model which can be visually inspected, and a plot of the sections would be helpful too.

@James Walters

Hi James,

Many thanks, but I don't want you to do anything tedious and dull. I'm not doing it manually -- I define the radial centre for say a rib, and Templot generates the nodes along it.

Preferably not a spreadsheet -- just a comma or space-separated list in ordinary .txt format would be great. X +/- from the centre-line of the chair, Y from the rail gauge-face. Z is constant for any given section, from the rail-top or the underside of the chair, whichever is easier. All in prototype inches for preference. It's symmetrical from the centre-line of the chair, so only one-half is actually needed:

X1 Y1
X2 Y2
X3 Y3

I can then convert that into Pascal code very easily.

I assumed that a CAD program could do this easily with a few clicks. When we were making tooling for turbine blades we got such lists of section co-ordinates from the customer along with the CAD drawing.

I can foresee the need to use parametric variables within the drawing for things such as rail section, loose-jaw slot etc, but it would be interesting to have a go.

Folks keep saying this, but in truth I haven't much idea what it means, and no idea at all how I would integrate it into the existing DXF/STL output. I think I would prefer to stick with my present methods, but if those following up on OpenTemplot2024 want to do things differently that's fine.

Extracting cross-sections from Solidworks is very simple, so if you can write a code to automate plotting the nodes then this could be a very good way forward.

Do you mean a bitmap image, such as a PNG file? Yes I can write edge-detection for that if it's easier. It needs to be at reasonably hi-res, and I would need to know the overall prototype size, or preferably dots per prototype inch if it's an odd shape.

If I've not misunderstood the task, would you like to suggest a particular chair for me to have a look at?

Well this GWR and BR(W) Ordinary chair for BS-95R rail has been in my garden for about 40 years. The outer jaw is almost identical to REA S1, so we can re-use that. But the inner jaw has 2 ribs instead of 1 to fit around the single chair bolt (this is a 2-bolt chair instead of REA 3-screw). Perhaps you could have a look at this inner jaw?

It's the chair on the left here in the drawing (R.F. is the running face, rail gauge-face, inner jaw to the right in the drawings, left in the photos):

The base dimensions are identical to the REA S1 which is already done, so this would be an easy one to start with. I can measure any dimensions you need directly from it.

On the right is the more traditional GWR Ordinary chair for GWR 00 rail. It's the one modellers probably mean when they ask for 2-bolt GWR chairs. It's an inch longer and half an inch narrower, but the inner jaw looks very similar to the BS-95R version -- we could probably use the same inner jaw for both. For the outer jaw we shall need something longer than the existing S1 outer. An interesting detail is that the inner corners on the base are 1/2" rad, and the outer corners are 1" rad. Someone is sure to pick that up if we don't get it right.

cheers,

Martin.

 

[James Walters]

That all makes perfect sense.
Solidworks will output the sections in pretty much any format, vector or bitmap. and the exact resolution can be specified. When I've got some sections we can figure-out the best type/resolution.
I'll ignore the outer jaw for now for simplicity (it can always be added later), and the loose-jaw socket.
Also, I'll leave the screw/bolt bosses plain unless you'd prefer otherwise.

Out of interest, would these have been screwed or bolted?

Ignore what I said about parametric modelling. It's useful to consider when creating a 3d model in CAD, but not relevant to the data capture you're after here. I'll have a play.

James

 

[James Walters]

That all makes perfect sense.
Solidworks will output the sections in pretty much any format, vector or bitmap. and the exact resolution can be specified. When I've got some sections we can figure-out the best type/resolution.
I'll ignore the outer jaw for now for simplicity (it can always be added later), and the loose-jaw socket.
Also, I'll leave the screw/bolt bosses plain unless you'd prefer otherwise.

Out of interest, would these have been screwed or bolted?

Ignore what I said about parametric modelling. It's useful to consider when creating a 3d model in CAD, but not relevant to the data capture you're after here. I'll have a play.

James

One other thing...Do you program in metric or imperial units?

 

[Martin Wynne]

One other thing...Do you program in metric or imperial units?

@James Walters

Yes.

For most things, including the chairs I use the prototype size in inches. Templot converts everything to model sizes in mm as needed. DXF output can be in model mm, or model inches, or can be scaled to whatever you want.

I'll ignore the outer jaw for now for simplicity (it can always be added later), and the loose-jaw socket.
Also, I'll leave the screw/bolt bosses plain unless you'd prefer otherwise.

Please just do the inner jaw (green):

That's the REA S1 chair, so we need a replacement inner jaw with two ribs to fit round a single screw/bolt. The ribs may need to be reduced a little to clear 00/EM wheel flanges.

The jaws are inserted in the DXF file as re-usable DXF blocks (different colours above), located and rotated to fit the rail. Likewise the bolts. If you put it all in one file I won't know how to split it up.

The S1 plinth height is 7/8" (0.875") above the underside of the chair. So to re-use the S1 base we need the inner jaw to sit flat on that surface.

Out of interest, would these have been screwed or bolted?

For the GWR and BR(W) on plain track they are bolted from below, with square nuts on top. When used in pointwork they are screwed from above, using square-head coach screws. The holes are parallel.

(For all REA chairs as above the holes are tapered and fitted with deep tapered ferrules for the chair screws. This makes them much more prominent than GWR.)

So we will needed replacement screw/bolt blocks for the above which are a bit smaller and lower. The rest of it can be unchanged from the S1. Maybe the plinth depth could be reduced a bit.

cheers,

Martin.

 

[Martin Wynne]

p.s. James,

Don't bother with the grip part of the jaw because it differs to match the rail section, and the existing grip code can be re-used. The jaw block just needs a vertical face against the rail foot.

cheers,

Martin.

 

[Steve_Cornford]

Hi Martin & James,
Can I suggest James starts with an REA S1 chair as a proof of concept, and you could then compare the figures that he produces with your version of the S1 chair.
That way the procedure can be verified, just like in an experiment.
Steve

 

[Phil O]

@James Walters

Hi James,

Many thanks, but I don't want you to do anything tedious and dull. I'm not doing it manually -- I define the radial centre for say a rib, and Templot generates the nodes along it.

Preferably not a spreadsheet -- just a comma or space-separated list in ordinary .txt format would be great. X +/- from the centre-line of the chair, Y from the rail gauge-face. Z is constant for any given section, from the rail-top or the underside of the chair, whichever is easier. All in prototype inches for preference. It's symmetrical from the centre-line of the chair, so only one-half is actually needed:

X1 Y1
X2 Y2
X3 Y3

I can then convert that into Pascal code very easily.

I assumed that a CAD program could do this easily with a few clicks. When we were making tooling for turbine blades we got such lists of section co-ordinates from the customer along with the CAD drawing.


Folks keep saying this, but in truth I haven't much idea what it means, and no idea at all how I would integrate it into the existing DXF/STL output. I think I would prefer to stick with my present methods, but if those following up on OpenTemplot2024 want to do things differently that's fine.


Do you mean a bitmap image, such as a PNG file? Yes I can write edge-detection for that if it's easier. It needs to be at reasonably hi-res, and I would need to know the overall prototype size, or preferably dots per prototype inch if it's an odd shape.


Well this GWR and BR(W) Ordinary chair for BS-95R rail has been in my garden for about 40 years. The outer jaw is almost identical to REA S1, so we can re-use that. But the inner jaw has 2 ribs instead of 1 to fit around the single chair bolt (this is a 2-bolt chair instead of REA 3-screw). Perhaps you could have a look at this inner jaw?

It's the chair on the left here in the drawing (R.F. is the running face, rail gauge-face, inner jaw to the right in the drawings, left in the photos):

View attachment 9770

The base dimensions are identical to the REA S1 which is already done, so this would be an easy one to start with. I can measure any dimensions you need directly from it.

On the right is the more traditional GWR Ordinary chair for GWR 00 rail. It's the one modellers probably mean when they ask for 2-bolt GWR chairs. It's an inch longer and half an inch narrower, but the inner jaw looks very similar to the BS-95R version -- we could probably use the same inner jaw for both. For the outer jaw we shall need something longer than the existing S1 outer. An interesting detail is that the inner corners on the base are 1/2" rad, and the outer corners are 1" rad. Someone is sure to pick that up if we don't get it right.

cheers,

Martin.

Martin and Others,

GW/ BR (W) 00 chairs are for 100lbs rail, this was originally used in tunnels, particularly the Severn Tunnel and areas likely to be wet, to prolong rail life, but over time cascaded down to lesser lines.

There were odd groups of 00 panels on the ESR, we were always on the lookout for compatibility between chairs and rail. At a quick glance it's difficult to pick out the differences between rails and chairs.

 

[Martin Wynne]

Martin and Others,

GW/ BR (W) 00 chairs are for 100lbs rail, this was originally used in tunnels, particularly the Severn Tunnel and areas likely to be wet, to prolong rail life, but over time cascaded down to lesser lines.

There were odd groups of 00 panels on the ESR, we were always on the lookout for compatibility between chairs and rail. At a quick glance it's difficult to pick out the differences between rails and chairs.

@Phil O

Hi Phil,

According to David Smith's book, the 00 rail was 97.5 lb/yd, and the 00 name ("nought-nought") refers to the year of its introduction -- 1900.

It's all a bit academic for models, because we have to use the rail sections available from the trade. Unless we 3D-print our own, of course.

cheers,

Martin.

 

[Phil G]

Hi Martin
re the GWR chair would a mirror of the two outer jaw ribs on say the rail C/L, thus creating two ribs onto the inner jaw be a close enough location of the two new ribs required? clearly the new inner jaw profile shape would then need to be replicated.
cheers
Phil,

 

[Martin Wynne]

Hi Martin
re the GWR chair would a mirror of the two outer jaw ribs on say the rail C/L, thus creating two ribs onto the inner jaw be a close enough location of the two new ribs required? clearly the new inner jaw profile shape would then need to be replicated.
cheers
Phil,

@Phil G

Hi Phil,

Not too sure what you mean there? The outer ribs can't be used on the inner jaw because they would hit the wheel flanges.

There are cross-sections for the jaws and the ribs on the REA S1 drawing -- page 4 at:

https://85a.uk/templot/club/index.php?resources/rea-bullhead-track-drawings.12/

Essentially we are looking to modify the inner jaw to have double ribs instead of a single rib.

cheers,

Martin.

 

[Phil G]

Essentially we are looking to modify the inner jaw to have double ribs instead of a single rib.

Hi Martin,
That is exactly what I was talking about, the mirror is a cad term for mirror image in this case a mirror image of the two ribs on the outer jaw. projected onto the inner jaw in terms of X and Y location only. Fully understand the Z height would be different hence my comment, "clearly the inner jaw profile (which on refection should have said inner jaw height profile) will need to be replicated."
If its all too hard I will step to one side and let James take it on.
cheers
Phil,

 

[Martin Wynne]

@Steve_Cornford

Hi Steve,

Thanks for finding that bug. It does illustrate how time-consuming it can be to test everything.

timbers only FDM should be switching the background shapes layer back on. I will get it fixed -- or maybe add yet another tick-box option to choose whether you want the brick assemblers included with FDM timbers.

The underlying issue is that I intended making use of the background shapes for the brick assemblers to be a temporary kludge. The assemblers need a separate function of their own, but I suspect it will be a while before I do anything about that.

Thanks again,

Martin.

@Steve_Cornford

Hi Steve,

Another tick-box for 245a:

Normally left on. Unticking it in 3D will over-ride other settings and switch the assemblers off. Ignored in 2D.

Thanks again for finding it.

Martin.

 

[James Walters]

@James Walters

Yes.

For most things, including the chairs I use the prototype size in inches. Templot converts everything to model sizes in mm as needed. DXF output can be in model mm, or model inches, or can be scaled to whatever you want.



Please just do the inner jaw (green):

View attachment 9772
That's the REA S1 chair, so we need a replacement inner jaw with two ribs to fit round a single screw/bolt. The ribs may need to be reduced a little to clear 00/EM wheel flanges.

The jaws are inserted in the DXF file as re-usable DXF blocks (different colours above), located and rotated to fit the rail. Likewise the bolts. If you put it all in one file I won't know how to split it up.

The S1 plinth height is 7/8" (0.875") above the underside of the chair. So to re-use the S1 base we need the inner jaw to sit flat on that surface.



For the GWR and BR(W) on plain track they are bolted from below, with square nuts on top. When used in pointwork they are screwed from above, using square-head coach screws. The holes are parallel.

(For all REA chairs as above the holes are tapered and fitted with deep tapered ferrules for the chair screws. This makes them much more prominent than GWR.)

So we will needed replacement screw/bolt blocks for the above which are a bit smaller and lower. The rest of it can be unchanged from the S1. Maybe the plinth depth could be reduced a bit.

cheers,

Martin.

Hi Marin,
I'm having a look at this now. Can you confirm the dimensions you'd like me to use as per the drawing below.
Best,
James

 

[James Walters]

Oh and also the distance between the rail seat and the base of the chair, if you don't mind. Ta.

 

[Martin Wynne]

@James Walters

Hi James,

Thanks for looking at this so quick!

What I need is the part of the inner jaw in the yellow box:

If we use the existing REA S1 chair base, the plinth dimension E is 7/8" (0.875").

Dimensions A, B, C, D are determined by the user's choice of model rail section, so the jaw co-ordinates are adjusted to fit when the jaw is generated.

For Templot everything is dimensioned from the rail gauge-face (white line), so your red centre-line wouldn't apply. Also all rails in plug track are vertical, and not inclined at 1:20 as shown in the prototype drawings. To the extent that the model bullhead rail-foot is not the same width as the rail-head, an offset is added accordingly when inserting the jaw block in the DXF.

All these adjustments are easily done in the code once I have the nodes for each horizontal section through the jaw. I can also easily stretch or compress the Z dimensions a little if we decide to change the plinth thickness. We can't get very close to the prototype base thickness because it makes the corners of the chair too fragile in the resin.

The thickness of the rail seat under the rail is 1.3/4" (1.75")*, but that doesn't play any part in the jaw design. If we are using the existing REA S1 as a basis for this GWR chair, the base, seat, outer jaw and key are already done. We shall need a smaller bolt boss and square bolt-head for pointwork -- here are 4 of the chairs we are wanting:

(screws!)

And with square nuts in plain track, outer jaw and inner:

This means that strictly speaking for GWR plain-track chairs we need 2 versions of the chair -- one with the end of the bolt protruding through the nut, and one with a flat top. But it would be easy to snip the nut flush.

As with the REA screws, the bolts will be inserted in the chairs with random nut rotations (you don't get that with injection-moulding -- and in Gauge 1 you can actually see it ).

*GWR differs slightly from REA, but we want these chairs to be inter-mixable in a timbering base, so we will stick to the REA dimension. For the GWR it's a bit academic because GWR chairs have a serrated base to resist gauge-spread, and a serrated housing for them is adzed into the timber surface.

Thanks again, it's great to have some help with this.

cheers,

Martn.

 

[James Walters]

That's perfect Martin, much easier too.
I'm drafting the various profiles in 2d first, and will do the whole chair, simply because because it might one-day be of some use. But I'll develop the yellow box area in 3d.
Thanks very much,
James

 

[Martin Wynne]

@James Walters

Hi James,

Many thanks.

I've brightened up this image to make it easier to see the chairs. They are the same as the chair from my garden, but it is easier to judge the proportions when viewed in situ in track.

An interesting detail in this photo is the square flat end on the check rail -- so unlike the chewed-off rail-ends you often see in models. With the loose jaws it is easy to do because no chamfer is need to assist in threading the chairs. Just a quick touch on the sanding disc is all that's needed.

cheers,

Martin.

 

[Hayfield]

Martin and James

If I can put my two pence in to the mix, my own thoughts especially in the smaller scales, having something rather than nothing is a compromise worth having.

Trying to cater for multi designs of the same chair might be what the odd modeler requires, but the majority who would like a 2 bolt chair would settle for a chair regardless if it has bolts or nuts

Thanks to both of you for pushing along the development

 

[Martin Wynne]

Martin and James

If I can put my two pence in to the mix, my own thoughts especially in the smaller scales, having something rather than nothing is a compromise worth having.

Trying to cater for multi designs of the same chair might be what the odd modeler requires, but the majority who would like a 2 bolt chair would settle for a chair regardless if it has bolts or nuts

Thanks to both of you for pushing along the development

@Hayfield

Hi John,

You are too late.

Whether they are nuts or bolt-heads will depend on whether they are generated from a plain track template or a pointwork template.

If nuts, it would be easy (but tedious) to snip off flush the protruding bit of bolt from the nut, converting it to a bolt-head.

It is important to note that the above will be the GWR 2-bolt Ordinary chairs for BS-95R rail, mostly used post-WW2. The traditional pre-group GWR chairs for GWR 1900 rail, which are longer and narrower, will have to wait until a full set of GWR chairs gets done.

The GWR 2-bolt Pit chairs for BS-95R are almost identical to the REA L1 bridge chairs, apart from having only 2 bolt-heads on diagonally opposite corners -- instead of 4 REA chair screws. So when I have the bolt-heads done for the above, it will be a quick fix to do the GWR Pit chairs too.

All the above will be interchangeable with REA in the sockets, so you will be able to choose which to use when assembling the plug track.

cheers,

Martin.

 

[Martin Wynne]

Hi Martin & James,
Can I suggest James starts with an REA S1 chair as a proof of concept, and you could then compare the figures that he produces with your version of the S1 chair.
That way the procedure can be verified, just like in an experiment.
Steve

@Steve_Cornford @James Walters

Hi Steve,

I don't think we should make James pass a test. I'm just very grateful that he is helping with this.

We could maybe send him some buttons?

cheers,

Martin.

 

[Martin Wynne]

@James Walters

Hi James,

Looking at your chair drawing and comparing it with the photos and the chair from my garden, the ribs on the inner jaw look much beefier than in the drawing. I think they should probably be similar to the rib on the REA chairs.

Here are the drawings for those. The S1J scan is easier to read than the S1 scan so I'm attaching both. The jaw detail is identical on both, only the width of the chair base differs. A cross-section through the jaw is shown -- we are looking for a whole stack of them at different levels.

Click the image twice to see the original scans, and drag them around.

cheers,

Martin.

 

[KHC1]

Anyone doing fishplates by any chance. Just so I can add a raft or two to a print.

Keith

If not any chance a dimensioned drawing so I could have a stab.

 

[Martin Wynne]

If not any chance a dimensioned drawing so I could have a stab.

@KHC1

Hi Keith,

Click the image twice to see it full-size:

I don't know of anyone doing them, it's likely that someone is and I have forgotten.

I'm intending to do them myself eventually, with a thin tab on the back to fit between the rail ends and provide insulation with a minimum rail gap. Note that if using drop-in construction with loose jaws, they can't be the slide-on H-section double fishplate type available from the trade. Each plate needs to be separate and applied to the side of the rail with adhesive after assembly. It would be purely cosmetic, the rail alignment is provided by the chairs. Two designs are needed -- one with nuts to go on the outside of the rail and one with dome-headed bolts to go on the inside of the rail. The tab on the back can be the full rail height, and half the rail-width, and say 0.2mm thick.

Fishplates are normally 18" long.

Fish-bolt hole centres are at 2.1/2" from the centre of the fishplate and 4.1/2" to the next. i.e. 4 bolts not evenly spaced -- centres 4.1/2" - 5" - 4.1/2".

The standard REA fish-bolt for BS-95R bullhead rail is:

15/16" dia. 4.3/4" long, Whitworth thread 9 tpi​

​

domed head is 1.3/4" dia. and 7/8" high.​

​

nut is 1.5/8" square and 1.1/8" thick.​

​

The head is normally on the gauge side of the rail, and the nut on the outside.

The above is the REA standard. Some GWR fish bolts use a different bolt design -- 1" dia. with a square head and hexagon nut.

Above are the REA drawings if you would like to have a stab at them. I'm sure the STL would be popular if you post it here.

cheers,

Martin.

 

[Matt Harris]

Here's my first go at it using FreeCAD:

Still need to work on the fasteners.

 

[James Walters]

@James Walters

Hi James,

Looking at your chair drawing and comparing it with the photos and the chair from my garden, the ribs on the inner jaw look much beefier than in the drawing. I think they should probably be similar to the rib on the REA chairs.

Here are the drawings for those. The S1J scan is easier to read than the S1 scan so I'm attaching both. The jaw detail is identical on both, only the width of the chair base differs. A cross-section through the jaw is shown -- we are looking for a whole stack of them at different levels.

Click the image twice to see the original scans, and drag them around.

View attachment 9801

View attachment 9802

cheers,

Martin.

Thank you Martin, sorry for this delayed reply. I've been in my studio today and the WiFi is iffy, so only just checking-in.
I agree that the jaw profile on the drawing I was using for reference are wrong. I was using a drawing for the early rail.
I've just emailed you some section views and will go back and make modifications to improve the jaw profile. It's a simple fix.
In the meantime if you take a look at the images I have sent through they hopefully give a guide as to what I can produce and whether or not they will be fit for your purposes.

All the best,

James

 

[Martin Wynne]

I've just emailed you some section views and will go back and make modifications to improve the jaw profile. It's a simple fix.
In the meantime if you take a look at the images I have sent through they hopefully give a guide as to what I can produce and whether or not they will be fit for your purposes.

@James Walters

Many thanks James.

Yes I can do edge detection on those images. Perhaps we need a few more, say at 1/4" layer spacing?

The ribs need to run out just before they reach the top of the stand, otherwise it makes it much trickier to insert the calculated grip part (green) behind the stand part (pink):

We need to get as much strength into this tiny component as possible -- the top of the jaw can be raised a bit to that effect (red lines).

The T-shaped dotted outline in the drawing is the prototype fouling bar. We don't need to be concerned with clearing that in a model, but we do need to bear in mind the deeper 00/EM wheel flanges. The above is roughly what I had in mind.

cheers,

Martin.

 

[Cransford]

@KHC1

Hi Keith,

Click the image twice to see it full-size:

View attachment 9805

I don't know of anyone doing them, it's likely that someone is.

I'm intending to do them myself eventually, with a thin tab on the back to fit between the rail ends and provide insulation with a minimum rail gap. Note that if using drop-in construction with loose jaws, they can't be the slide-on H-section double fishplate type available from the trade. Each plate needs to be separate and applied to the side of the rail with adhesive after assembly. It would be purely cosmetic, the rail alignment is provided by the chairs. Two designs are needed -- one with nuts to go on the outside of the rail and one with dome-headed bolts to go on the inside of the rail. The tab on the back can be the full rail height, and half the rail-width, and say 0.2mm thick.

Fishplates are normally 18" long.

Fish-bolt hole centres are at 2.1/2" from the centre of the fishplate and 4.1/2" to the next. i.e. 4 bolts not evenly spaced -- centres 4.1/2" - 5" - 4.1/2".

The standard REA fish-bolt for BS-95R bullhead rail is:

15/16" dia. 4.3/4" long, Whitworth topic 9 tpi​

​

domed head is 1.3/4" dia. and 7/8" high.​

​

nut is 1.5/8" square and 1.1/8" thick.​

The head is normally on the gauge side of the rail, and the nut on the outside.

The above is the REA standard. Some GWR fish bolts use a different bolt design -- 1" dia. with a square head and hexagon nut.

Above is the REA drawing if you would like to have a stab at them. I'm sure the STL would be popular if you post it here.

cheers,

Martin

@KHC1

Hi Keith,

Click the image twice to see it full-size:

View attachment 9805

I don't know of anyone doing them, it's likely that someone is.

I'm intending to do them myself eventually, with a thin tab on the back to fit between the rail ends and provide insulation with a minimum rail gap. Note that if using drop-in construction with loose jaws, they can't be the slide-on H-section double fishplate type available from the trade. Each plate needs to be separate and applied to the side of the rail with adhesive after assembly. It would be purely cosmetic, the rail alignment is provided by the chairs. Two designs are needed -- one with nuts to go on the outside of the rail and one with dome-headed bolts to go on the inside of the rail. The tab on the back can be the full rail height, and half the rail-width, and say 0.2mm thick.

Fishplates are normally 18" long.

Fish-bolt hole centres are at 2.1/2" from the centre of the fishplate and 4.1/2" to the next. i.e. 4 bolts not evenly spaced -- centres 4.1/2" - 5" - 4.1/2".

The standard REA fish-bolt for BS-95R bullhead rail is:

15/16" dia. 4.3/4" long, Whitworth thread 9 tpi​

​

domed head is 1.3/4" dia. and 7/8" high.​

​

nut is 1.5/8" square and 1.1/8" thick.​

The head is normally on the gauge side of the rail, and the nut on the outside.

The above is the REA standard. Some GWR fish bolts use a different bolt design -- 1" dia. with a square head and hexagon nut.

Above is the REA drawing if you would like to have a stab at them. I'm sure the STL would be popular if you post it here.

cheers,

Martin.

Matt/Martin,

I've previously had a go at fishplates based on the current design (RE/PW drawings) and produced prints of both the unskirted and skirted types. I did post details but it was a while ago and didn't fancy trawling back! Pictures of skirted types;

And, yes, I know I got one section of rail upside down in picture!

Having produced a 'mark one' version, I realised some changes to fishing surfaces and fit was required to allow for difference between real and model rail.

STL files for for types - 54253 (unskirted) and 54293 (skirted) also attached.

I introduced peg to raise plates off print bed and to make handling a little easier.

Cheers,

Paul

 

[KHC1]

Thanks to all for the fishplates. I will add it to my list of cad projects and see what I can do.

Keith

 

[KHC1]

Ok first attempt at the fishplate. Drawn using Designspark Mechanical in Imperial dimensions.

 

[KHC1]

Further development of fishplates showing outer and inner faces.
Keith

 

[Cransford]

Further development of fishplates showing outer and inner faces.
KeithView attachment 9913

Keith,

Looking good thus far - however!

If my memory serves the various constituent company designs were adapted/adopted to become "SRE" (Standard Railway Equipment I think) series drawings. These then went on to become the current "RE/PW" series drawings. Although I don't recall seing an SRE drawing for fishplates, to my knowledge the unskirted bull head fishplate had "tear drop" shaped holes and the bolt was a 15/16" diameter Pan/Pear/Square. This design allowed one man to tighten fishbolts. As you've shown with Square/Round/Square bolts you'd need two men to tighten as there is no headlock for the bolt head (don't forget the old patrolman was usually working on his own).

Unskirted fishplates were forged whereas the skirted type was a much more cost effective rolled section cut to length and drilled and were used with head lock washerplates.

I'll happily stand corrected but think there'd be a need for both bolt head plates!

Cheers,

Paul